230Th in the eastern North Atlantic:: The importance of water mass ventilation in the balance of 230Th

The Th-230 distribution in the water column on a north-south transect in the eastern North Atlantic was measured using thermal ionization mass spectrometry (TIMS). Water samples at three locations (L1, 33 degrees N 21.6 degrees W; L2: 47.4 degrees N 19.5W; and L3, 54.4 degrees N 21.1 degrees W) were filtered and the dissolved and particulate fractions were analysed for their Th-230 and Th-232 contents. In the upper water column (<1000 m water depth), the total Th-230(ex) distribution at the three locations could be explained by a reversible scavenging model. In the deep water column, however, Th-230(ex) concentrations decrease from south to north with relatively low and constant concentrations at stations L2 and L3. The major water mass below 2000 m is North East Atlantic Deep Water (NEADW), filling the deep water column at L1. The water column at L2 and L3 is influenced by Labrador Sea Water (LSW) down to 2000 m, whereas at L3, the deep water is strongly influenced by Iceland-Scotland Overflow Water (ISOW). The anomalous distribution of Th-230(ex) in the deep water column of stations L2 and L3 could be described after a ventilation term was added to the equations of the reversible scavenging model. For L3, a ventilation time (3-12 years) was found comparable to results derived from other tracers. However, the ventilation of the deep water (>2000 m) at L2 suggested by our model (8-25 years) is probably too rapid in comparison with previous findings based on chlorofluorocarbon data. Enhanced scavenging of Th-230 would be an alternative explanation for the low concentration in NEADW at L2. With the scavenging-mixing model, we calculated an advective export flux of Th-230 Of up to 50% of the production in the water column in the eastern North Atlantic. Advective loss of Th-230 has to be taken into account when using the Th-230 AUX into sediments for paleoceanographic studies in ocean basins where deep water residence time is comparable to the Th-230 scavenging residence time. (C) 1998 Elsevier Science B.V. All rights reserved.